How does retirement affect one's health? Because individuals are living longer and need to remain economically active until later ages to ensure that old age insurance systems are sustainable, the relationship between health and work has become a primary policy concern and has received much attention in the recent economic literature. Summarizing, this literature finds that retirement does not have a negative impact on health and may lead to health improvements, although the size and significance of this finding varies between studies and countries. While these studies examine the causal link between retirement and health, none of them go further to discover the mechanisms behind the estimated effects. Most of the existing literature that examines the relationship between retirement, health behaviors, and health investments does not get at the causal relationship. This is because these studies often ignore the fact that retirement is an individual choice that may be caused by declining health. While the health effect of retirement is interesting in its own right, it is currently impossible to draw policy conclusions about how to impact the causal relationship. This study will further our understanding of the causal pathways to maintaining better health in retirement. We will distinguish between behaviors individuals have direct control over (physical activity, social networks, eating, drinking, smoking) and those at least partially governed by the health care system (visits to the doctor, preventative care). We will make use of panel data and/or instrumental variable methods, exploiting information about statutory retirement ages, to assess how retirement causally affects health-related investments and behaviors. We propose to use panel data from three recently available, harmonized, longitudinal studies on middle-aged and elderly adults in three different settings: eight waves of the Health and Retirement Study (HRS) (United States);three waves of the English Longitudinal Study of Ageing (ELSA) (England);and two waves of the Survey of Health Aging and Retirement in Europe (SHARE) (continental Europe). This study will examine the following specific questions: 1. Which individual health-related behaviors are changing during retirement? 2. What role does retirement play in health care utilization? 3. Which pathways are the most important in explaining the effect of retirement on health and the differences found between countries? PUBLIC HEALTH RELEVANCE: While the causal health effect of retirement is interesting in its own right, it is currently impossible to draw policy conclusions about how policymakers or individuals can impact or change the causal relationship. Knowing what drives this causal relationship is important to determine what behaviors and health care utilization patterns may yield better health for the long-term. Also, understanding the underlying causal mechanisms can help inform policymakers about where they should place the emphasis of health care policy: on public health information, on behavior-changing programs, or on increasing access to the health care system in general.